SCIENCE & TECHNOL09YJ
November, 1973
Vol. 22, No. 11
UNTIL DECEMBER 31, 1973
CD
STARTING JANUARY 1, 1974
computers
and people
WORKING TOGETHER
Cooperative Facilities to Obtain the Advantages of Computers - M. J. Cerullo
Control in Time-Sharing Systems - F.
C.
CastilloComputer Art: The Search Beyond Manipulation -
G. C.
HertleinComputers in Science Fiction - M. Ascher
Strategy and Action on World Trade - J. H. Binger Virtue, in Spite of Erroneous Conceptions - J. P. Frankel
Nixon and the Mafia - Conclusion - J. Gerth
~
2. 1721
16 511.3Here is the start of the most famous article that we ever published - excerpted from the May, 1970, issue of Computers and Automation. If you would like to read this article, and look at the eleven photographs it contains, send us $2 (prepayment is necessary). This issue is RETURNABLE IN 7 DAYS FOR FULL REFUND (IF IN SALABLE CONDITION), How can you lose?
Computers and Automation, 815 Washington St., Newtonville, Mass. 02160
THE ASSASSINATION OF PRESIDENT JOHN F. KENNEDY:
THE APPLICATION OF COMPUTERS
TO THE PHOTOGRAPHIC EVIDENCE
byPart 1. Introduction
Who Assassinated President Kennedy?
On November 22, 1963, in Dallas, Texas, President John F. Kennedy, while riding in an open limousine through Dealey Plaza and waving to the surrounding crowds, was shot to death. Lee Harvey Oswald, an ex-Marine, and former visitor to the Soviet Union, was arrested that afternoon in a movie theatre in another section of Dallas; that night he was charged with shooting President Kennedy from the sixth floor easternmost window of the Texas School Book Deposi- tory Building overlooking Dealey Plaza. This act Oswald denied steadily through two days of question- ing (no record of questions and answers was ever preserved). Two days later while Oswald was being transferred from one jail to another, he was shot by Jack Ruby, a Dallas night-club owner, in the basement of the Dallas police station, while mil- lions of Americans watched on television. The com- mission of investigation, appointed by President Lyndon B. Johnson, and headed by Chief Justice Earl Warren of the U. S. Supreme Court, published its report in September 1964, and concluded that Oswald was the sole assassin and that there was no conspir- acy.
In view of the authority of the Warren Commis- sion, that conclusion was accepted by many Americans for a long time. But the conclusion cannot be con- sidered true by any person who carefully considers the crucial evidence - such as ,the physics of the shooting, the timing of a number of events, and other important and undeniable facts. In other words, Oswald was not the sole assassin, and there was a conspiracy.
This article will develop that thesis, prove it to be true on the basis of substantial, conclusive evidence, and in particular some analysis of the photographic evidence.
There was in fact a conspiracy. Oswald played a role in the conspiracy, although there is con- clusive evidence that on November 22, 1963, he did no shooting at President Kennedy, and that, just as he claimed when he was in the Dallas jail, he was a "patsy." At least three gunmen (and proba- bly four) - none of whom were in the sixth floor easternmost window of the Texas School Book Depos- itory building where the Warren Commission placed Oswald - fired a total of six shots at President Kennedy.
One of these shots missed entirely; one hit Governor John B. Connally, Jr. of Texas, riding with Kennedy; and four hit President Kennedy, one
in his throat, one in his back, and two in his 2
Richard E. Sprague Hartsdale, New York
head. (The bulk of the undeniable evidence for these statements about the shots consists of:
(a) the physics of the motions of Kennedy and Con- nally shown in some 60 frames of the famous film by Abraham Zapruder; (b) the locations of the in- juries in Kennedy and in Connally; and (c) more than 100 pictures, consisting of more than 30 still photographs and more than 70 frames of mo- vies.)
More than 50 persons were involved in the con- spiracy at the time of firing the shots. These persons included members of the Dallas police force (but not all of the Dallas police - and that ac- {continued in the May 1970 issue of Computers and Automation}
Contents Parts 1 Introduction
2 The Photographic Evidence 3 The Application of Computers to
the Photographic Evidence 4 Appendices:
1 2 3 4 5 6, 7
8, 9 10 11
1 2
1 2 3 4 5
Acknowledgements and Notices Epilogue
Bibliography
Figures
Helicopter View of Dealey Plaza Policemen and "Tramps"
"Tramps"
Policemen and "Tramps"
Policemen and "Tramps"
6th Floor Easternmost Window of Texas School Book Depository Building
the Kennedy About the Time of the First
Shot
Kennedy After the First Three Shots and Before the Fatal Shot
The Radio Communicator Charts Spatial Chart
Schematic Timing Chart Tables Index to Spatial Chart
Photographs Acquired by FBI and Unavailable
Main List of Photographs
Preliminary List of Computer Codes Preliminary Coding Sheet for
Computer-Assisted Analysis
Page 2.0 34 56 58 59 60 37 38 39 40 41 42 43 44, 45 33
48, 49 51 46 50 52 57 58
COMPUTERS and AUTOMATION for November. 1973
THE PURSUIT OF IMPORTANT TRUTH
The magazine Computers and Automation has for more than three years followed an unusual publication policy:
- The pursuit of truth in input, output, and pro- cessing, for the benefit of people,
and an unusual belief:
- That computers are too important to be left to computer experts and must be integrated into a socially responsible profession of information engineering.
Where this policy has operated most is in publishing information, articles, and reports on subjects which a great many liberal and progressive newspapers and peri- odicals have left unexplored or unmentioned:
- The political conspiracies which have led to the assassination of President John F. Kennedy, Sena- tor Robert Kennedy, Martin Luther King, and others - and their coverups
- The conspiracies, coverups, and lies in connection with the pursuit. of war in I ndochina and dictator- ship by the Saigon regime
- The connections of President Richard M. Nixon with organized crime and the Mafia
- The Watergate crimes
If you believe in the value of truthful, frank reporting on the most important topics for the welfare of the
people of the United States today, we urge you to sub- scribe to our magazine, and buy our back copies (almost everyone is in print). Please help us pursue the important truth and report on it, by buying our products.
Edmund C. Berkeley, Editor
Unsettling, Disturbing, Critical ...
Computers and Automation, established 1951 and therefore the oldest magazine inthe field of com- puters and data processing, believes that the pro- fession of information engineer includes not only competence in handling information using computers and other means, but also a broad responsibility, in a professional and engineering sense, for:
--- The reliability and social significance of pertinent input data;
-- The social value and truth of the output results.
In the same way, a bridge engi neer takes a pro- fessional responsibility for the reliability and significance of the data he uses, and the safety and efficiency of the bridge he builds, for human beings to risk their lives on.
Accordingly, Computers and Automation publishes from time to time articles and other information related to socially useful input and output of data systems in a broad sense. To this end we seek to publish what is unsettling, disturbing, critical --- but productive of thought and an improved and safer "house" for all humanity, an earth in which our children and later generations may have a fu- ture, instead of facing extinction.
The professional information engineer needs to relate his engineering to the most important and most serious problems in the world today: war, nuclear weapons, pollution, the population explo- sion, and many more.
- - - (may be copied on any piece of paper) - - - - TO: Computers and Automation (Computers and People, starting January 1, 1974)
815 Washington St., Newtonville, Mass. 02160, U.S.A.
YES, please start my subscription to your magazine -
) Without the "Computer Directory": I enclose ( ) U.S.A., $11.50 ( ) Canada, $12.50 ) With the "·Computer Directory": I enclose ( ) U.S.A., $23.50 ) Canada, $24.50 ( ) Please bill my organization.
) Please send me information about the important articles in back copies (usual cost, $2 each)
) Foreign, $17.50 ) Forei~n, $32.50
Name: _______________________________________________________ Title ______________________________ __
Organization: ____________________________________________________________________________________ ___
Address: ______________________________________________________________________________________ __
Signature: ________________________________________ ___ Purchase Order No. ____________________________ _
COMPUTERS and AUTOMATION for November, 1973 3
Vol. 22, No. 11 November, 1973
Editor Assistant
Editors
Software Editor Advertising
Director Contributing
Editors
Advisory Committee
Editorial Offices
Advertising Contact
Edmund C. Berkeley Barbara L. Chaffee Linda Ladd Lovett Neil D. Macdonald Stewart B. Nelson
Edmund C. Berkeley
John Bennett Moses M. Berlin Andrew D. Booth John W. Carr III Ned Chapin Leslie Mezei
Bernhard W. Romberg Ted Schoeters Richard E. Sprague Ed Burnett James J. Cryan Bernard Quint
Berkeley Enterprises, Inc.
815 Washington St.
Newtonville, Mass. 02160 617-332-5453
The Publisher
Berkeley Enterprises, Inc.
815 Washington St.
Newtonville, Mass. 02160 617-332-5453
"Computers and Automation" is pub- lished monthly, 12 issues per year, at 815 Washington St •• Newtonville, Mass. 02160, by Berkeley Enterprises, Inc. Printed in U.S.A. Second Class Postage paid at Boston, Mass., and additional mailing points, •
Subscription rates: United States, $11.50 for one year, $22.00 for two years. Canada:
add $1 a year; foreign, add $6 a year.
NOTE: The above rates do not include our publication "The Computer Directory and Buyer,s' Guide". If you elect to re- ceive "The Computer Directory and Buyers' Guide", please add $12.00 per year to your subscription rate in U.S. and Canada, and
$15.00 per year elsewhere.
Please address all mail to: Berkeley Enterprises, Inc., 815 Washington St., New- tonville, Mass. 02160.
Postmaster: Please send all forms 3579 to Berkeley Enterprises, Inc., 815 Washing- ton St., Newtonville, Mass. 02160.
© Copyright 1973, by Berkeley Enter- prises, Inc.
Change of address: If your address changes, please send us both your new address and your old address (as it ap- pears on the magazine address imprint), and allow three weeks for the change to be made.
4
UNTIL DECEMBER 31, 1973
computers
and automation
STARTING JANUARY 1, 1974
computers
and people
The Computer Industry
8 Cooperative Facilities to Obtain the Advantages of Computers
by Michael J. Cerullo, State University of New York, Albany, N.Y.
[T A]
How to plan, before operations begin, a cooperative service bureau that will provide its sponsors with all the advantages of electronic data processing, plus the advantages of greatly reduced expenses.
10 Control in Time-Sharing Systems [T A]
by Fermin Caro del Castillo, Fort Worth, Texas
How time-shared computer systems should be controlled, made secure, and protected against incursions and hazards.
14 Alienation and the Systems Analyst [T A]
by Alan E. Brill, The Chase Manhattan Bank, New York, N.Y.
How systems analysts are often looked upon as in a
"computer department" and outside of the firm - and what might be done to correct this practice.
Computers and Art
18 Computer Art: The Search Beyond Manipulation [T A]
by Grace C. Hertlein, California State University-Chico, Chico, Calif.
How computer art is ranging through variation in patterns, variations in design, and varying philosophies of art - and where it may go.
Computers and the Future
20 Computers in Science Fiction - II [NT A]
by Marcia Ascher, Professor of Mathematics, Ithaca College, Ithaca, N.Y.
A survey of some two dozen themes of importance in the real world that are reflected in science fiction when astute writers explore the significance of computers to human beings.
6 The Understanding of Natural Language by Computers [NT E) by Edmund C. Berkeley, Editor, Computers and Automation
Some rather convincing evidence suggests that before long some large areas of ordinary natural language will be un- derstood by computers.
World Affairs and Social Policy
15 Strategy and Action on World Trade [NT A]
by James H. Binger, Chairman, Honeywell Inc., Minneapolis, Minn.
Why it is important for the well-being of the people of the United States to a.id the international division of
labor and production: - a discussion by the head of a multinational company which is a computer company.
COMPUTERS and AUTOMATION for November, 1973
The magazine of the design, applications, and implications of information processing systems - and the pursuit of truth in input, output, and processing, for the benefit of people.
World Affairs and Social Policy (continued)
33 Virtue, in Spite of Erroneous Conceptions [NT A]
by J. P. Frankel, Dean of the Faculty, Harvey Mudd College, Claremont, Calif.
Which projects and problems should scientists work on?
and which projects and problems should receive support by the government and which from other sources?
The Profession of Information Engineer and the Pursuit of Truth
3 Unsettling, Disturbing, Critical [NT F]
Statement of policy by Computers and Automation 3 The Pursuit of I mportant Truth
by Edmund C. Berkeley, Editor
Four kinds of truth that are regularly unexplored and unmentioned .
[NT F]
36 Nixon and the Mafia - Conclusion [NT A]
by Jeff Gerth, SunDance Magazine, San Francisco, Calif.
The many connections of President Richard M. Nixon with organized crime, scandal, etc.
26 Burying Facts and Rewriting History - II [NT A]
by Edmund C. Berkeley, Editor, Computers and Automation Taken together the information published May 1970 to
November 1973 in Computers and Automation effectively destroys a large segment of the beliefs, the rewritten history, that the establishment in the United States has arranged for the people in the United States to believe.
28 Political Assassinations in the United States [NT R]
Inventory of 41 articles published in Computers and Automation May 1970 to October 1973 on the assassi- nations of President John F. Kennedy, Senator Robert Kennedy, Reverend Martin Luther King, and other polit- ically important persons in the United States: titles, authors, and summaries.
27 The Watergate Crimes [NT R]
Inventory of 12 articles published in Computers and Automation August 1972 to September 1973 on the burglarizing of the Democratic National Committee offices in the Watergate Building, Washington, D.C., June 17, 1972, and the ramifications: titles, authors, and summaries.
Computers, Puzzles, and Games
35 Numbles
[T C]
by Neil Macqonald
Corrections
For changes in liThe Path to Championship Chess by Computer" by ProfeSsor Donald Michie published in the January 1973 issue of Computers and Automation, see page 23 of this issue, or page 24 of the July issue.
For changes in the reprinting and the indexing of "Communication - Three Way: Chimpanzee, Man, Computer",published in the July issue, see page 32 of this issue.
COMPUTERS and AUTOMATION for November, 1973
Front Cover Picture
Three eighth graders - each from a different continent - are working together to solve complex mathe- matics problems. They are at the new United Nations International School, and are Laurence Ling May, a Chinese-Thai American; Catharina Nilson of Stockholm, Sweden; and Arun Alagappan of I ndia. The minicomputer system is a gift from Digital Equipment Corp. For more information, see page 43 of the July 1973 issue.
Departments
42 Across the Editor's Desk - Computing and Data Processing Newsletter 50 Advertising Index
50 Calendar of Coming Events 32 Classified Advertisement 48 Monthly Computer Census 46
47 23
Key
[A]
[C]
[E]
[F]
[NT]
[R]
[T]
New Contracts New Installations Statement of Ownership
- Article
Monthly Column Editorial
Forum Not Technical Reference - Technical
NOTICE
*0 ON YOUR ADDRESS IMPRINT MEANS THAT YOUR SUBSCRIP- TION INCLUDES THE COMPUTER DIRECTORY. *N MEANS THAT YOUR PRESENT SUBSCRIPTION DOES NOT INCLUDE THE COM- PUTER DIRECTORY.
5
EDITORIAL
The Understanding of Natural Language by Computers
Proposition: Computers are making long strides towards understanding natural language as used by human beings, and will eventually understand such language as well as many human beings do.
What do we mean by: computers? understanding lan- guage? natural language? And what is the evidence for this proposition?
The word "computers" here refers to powerful comput- ers that have appropriate programs written by human be- ings plus the programmed capacity to improve their pro- grams by using experience. An example of such improve- ment is Dr. A. L. Samuels' famous checker-playing pro- gram, which can learn from experience, and which plays far better checkers than Dr. Samuels himself can play.
The experience may consist of the following at least:
Answers from human beings to questions posed by the computer;
Differences between computed results and a priori specified results;
Signals from the environment, such as instrument readings;
Information obtained by "looks" at the environment, as for example recognition of the character A, as in optical character recognition;
The interpretation of words, as with FORTRAN ex- pressions.
Probably there are even more categories of experience which a powerful computer program can use to modify it- self to become even a better program.
In regard to "understanding language", there are over 500 languages which computers have been programmed to understand, when written precisely according to stated rules: among them, BASIC, FORTRAN, COBOL, many kinds of machine language, etc.
"Natural language" is the ordinary language used by hu- man beings, subject to the requirement that for a computer implicit understandings must be stated. For example, when a speaker exclaims "Ouch!", a great deal of the meaning comes from the listener's observations of the speaker. But a computer, like a blind man, cannot "see" the situation, and must be told more than the ordinary listener.
With some definitions taken care of, let us consider evi- dence for the main proposition stated above. This evidence includes the information published in several recent ar- ticles in Computers and Automation.
The article '''Do What I Mean' - The Programmer's As- sistant", by Warren Teitelman, in the Apri11972 issue, described a "programmer's assistant", called "DWIM"; this was a "front end" or preprocessor to an interactive pro- gram for operating with LISP expressions. The front end was tolerant of the programmer's mistakes in typing, in low- level logic, etc.; it would catch misspellings, failures to have a balancing number of left and right parentheses, etc. If a
6
mistake occurred, it would inquire of the human program- mer what he meant, suggesting the correctly spelled alter- native, and so on. In this way the human programmer was relieved of much of the burden of expressing himself ex- actly and correctly the first time, and his efficiency in using the LISP interactive program greatly increased.
The three articles "Computer Programming Using Natu- ral Language" by Edmund C. Berkeley, Andy Langer, and Casper Otten, in the June, July, and August 1973 issues, demonstrated the understanding by a computer program called GENIE of at least some sets of instructions in ordi- nary natural language. The vocabulary though small was free, and there was unlimited freedom in putting the words together, with about a 90% chance of complete under- standing. The computer program that did this could be called 15% GENIE, because, as the authors emphasized, it was still in an early stage of development.
The article "Latest Computers See, Hear, Speak, and Sing - and May Outthink Man" by David Brand, in the October 1973 issue, enumerated many instances of com- puters (and robots equipped with computers) which could deal with concepts expressed in natural language. One of the programs mentioned could for example decipher natu- rallanguage commands dealing with the stacking of blocks of various shapes, sizes, and colors.
The main tasks for understanding natural language ap- pear to be the following:
Recognition of the framework of a sentence;
Recognition of the common meaning of groups of synonyms, the collection of words that "say the same thing";
Knowledge of context: the context is regularly specified to the computer, so that it does not have to deduce the context from "what is being said";
Knowledge of a limited vocabulary consisting of per- haps 300 to 500 words - knowledge in the sense that the computer program can attach meaning to the words either by themselves or in phrases, as for example it may attach the meaning of doing something three times to the numeral 3;
Capacity to accept variation in the way something is stated by a human programmer.
The three articles referred to above show instances of the achievement of all of these tasks separately. Combin- ing all the achievements should not be too hard, especially since children as young as four years old show complete capacities to listen to many kinds of natural language, un- derstand, and respond.
C=~tAC.~
Edmund C. Berkeley Editor
COMPUTERS and AUTOMA liON for November. 1973
The Notebook on
COMMON SENSEI ELEMENTARY AND ADVANCED
is devoted to development, exposition, and illustration of what may be the most important of all fields of knowledge:
WHAT IS GENERAllY TRUE AND IMPORTANT
+ +
JUDGEMENT ANDMATURITY
+
SCIENCE IN GENERAL
TECHNIQUES FOR AVOIDING
MISTAKES
+
PURPOSES:
to help you avoid pitfalls
+
to prevent mistakes before they happen to display new paths around old obstacles to point out new solutions to old problems to stimulate your resourcefulness
to increase your accomplishments to improve your capacities to help you solve problems
to give you more tools to think with
8REASONS TO BE INTERESTED IN THE FIELD OF COMMON SENSE, WISDOM, AND GENERAL SCIENCE COMPUTERS are important -
But the computer field is over 25 years old. Here is a new field where you can get in on the ground floor to make your mark.
MATHEMATICS is important -
But this field is more important than mathematics, because common sense, wisdom, and general science have more
+
Topic:
TH E SYST EMA TI C
PREVENTION OF MISTAKES Already Published
Preventing Mistakes from:
Failure to Understand Forgetting
Unforeseen Hazards Placidity
To Come
Preventing Mistakes from:
Bias Camouflage Interpretation Distraction Gullibility Failure to Observe Failure to Inspect Prejudice
+
+
TECHNIQUES FOR SOLVING PROBLEMS
Topic:
SYSTEMATIC EXAMINATION OF GENERAL CONCEPTS
Already Published The Concept of:
Expert Rationalizing Feedback Model Black Box Evolution Niche To Come
Strategy Understanding Teachable Moment Indeterminacy System
Operational Definition
+
applications.
WISDOM is important -
This field can be reasonably called tIthe engineering of wisdom".
.. - - - (may be copied on any piece of paper) - - - .. - - - - •• - - , To: Computers and Automation
815 Washington St., Newtonville, Mass. 02160
COMMON SENSE is important _ ' ( ) Yes, please enter my subscription to The Notebook on Common This field includes the systematic study and development of ' Sense, Elementary and Advanced at $12 a year (24 issues), plus
common sense. extras. I understand that you always begin at the beginning
SCI ENCE is important _ and so I shall not miss any issues.
This field includes what is common to all the sciences, what : ( ) Please send me as free premiums for subscribing:
is generally true and important in the sciences. ' 1. Right Answers - A Short Guide to Obtaining Them 4. Strategy in Chess
MISTAKES are costly and to be AVOIDED - , 2. The Empty Column 5. The Barrels and the Elephant This field includes the systematic study of the prevention of ' 3. The Golden Trumpets of Yap Yap 6. The Argument of the Beard
mistakes. ( ) I enclose $ ) Please bill my organization
MONEY is important - RETURNABLE IN 7 DAYS FOR FULL REFUND IF NOT SATISFACTORY
The systematic prevention of mistakes in your organization HOW CAN YOU LOSE?
might save 10 to 20% of its expenses per year.
OPPORTUNITY is important -
If you enter or renew your subscription to both Computers and Automation and the Notebook on Common Sense at the same time, direct to us, - you may take off $2.00 per year from the total cost.
COMPUTERS and AUTOMATION for November. 1973
Name _ _ _ _ _ _ _ _ ~ _ _ _ _ _ Title _ _ _ _ _ _ _ _ Organization--'---_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Address (including zip) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Signature _ _ _ _ _ _ _ _ _ _ _ Purchase Order No. _ _ _ _
7
Cooperative Facilities to Obtain the Advantages of Computers
Michael J. Cerullo
Asst. Professor of Accounting State Univ. of New York at Albany 1400 Washington Ave.
Albany, N. Y. 12222
"Because of general dissatisfaction, 30% of the surveyed service bureau clients planned to discontinue using service bureaus."
Computer Use
A recent study of 2,500 companies conducted by the Research Institute of America, revealed that 55%
of the firms regularly use computers in conducting their business. Further breakdown of the study shows that an average of 32% own or lease their own computers, and 23% use an outside service bureau for processing data. l
While a majority of the companies surveyed do use computers in some form, a significant 45%.of the re- spondents do not use any type of computerized data processing service. Those companies cited the fol- lowing reasons for not using computers: 2
'Operation too small Too costly
Looked into and tabled for the present Present methods satisfactory
plan to install a computer within a year
Plan to begin using a service bureau within a year
.Other
Total (multiple answers) Service Bureau Use
Per Cent 50 35 31 21 8 8
166%
In addition to the non-users, another recent study revealed that most service bureau clients are not effectively using their service bureau and, as a result, are receiving few, if any, of the advan- tages q! electronic data processing (EDP). Most of th~surveyed service bureau clients were: 3
8
1. Receiving routine services which do not save them money.
2. Not planning to expand into more sophi sticated, higher-payoff applications.
3. Not receiving indirect benefits and savings, such as improved information for decision making.
4. Not satisfied with current service&.
The study also disclosed that because of the gen- eral dissatisfaction, 30% of the surveyed service bureau clients planned to discontinue using service bureaus and purchase in-house computers. 4
Cooperative Service Bureaus
Both non-users and dissatisfied service bureau clients, therefore, offer a ready market for a newer approach to obtaining the advantages of EDP - the formation of a cooperative service bureau. A coop- erative service bureau consists of several sponsor- ing firms who jointly own a computer and share in its operating costs at a much lower expense than in- dividual ownership of a computer. If carefully planned before operations begin, a cooperative ser~
vice bureau will provide its sponsors with all the advantages of EDP plus many other advantages unique to jointly-shared facilities and personnel.
Among these additional advantages are:
1. The sponsors would purchase a computer tai- lored to their specific needs.
2. They would have exclusive use of the computer.
3. Their computer would be available for use at any time.
4. They could share the costs of any computer programs developed.
5. They could share the costs of any package or canned computer programs purchased.
6. They would have available for their exclusive use a staff of data processing experts and a computer - at a fraction of the cost of individual ownership.
Factors Insuring the Success of the Cooperative The first step in forming a successful coopera- tive service bureau is to enroll an appropriate num- ber of comapnies. Ideally the cooperative should be limited to five or six sponsoring firms. A larger number may create unmanageable problems involving operations, communications. control, scheduling of computer time, and so forth. A smaller number may make the venture economically unfeasible.
COMPUTERS and AUTOMATION for November. 1973
Lack of· Interest
At the present time, widespread lack of interest in joining a cooperative is more of a problem to po- tential organizers than too much demand. The author contacted a number of public accounting firms about the feasibility of forming cooperatives and found that three firms who had already made such an at- tempt could not interest enough firms to join. Their comments were:
We tried to organize one several years ago;
the firms contacted were not interested.
We would like to explore this possibility but we can't get anyone else interested.
At present there appears to be a tragic lack of interest locally by other CPA firms.
Unquestionably, one reason for the lack of inter- est is a misunderstanding of the true nature of a cooperative. Interested companies must therefore be prepared to sell the idea to others. As more pub- lished material becomes available in the field, this problem should be alleviated.
Confidential Information
A second major reason for lack of interest is that companies fear for the control and security of their confidential or sensitive information. Such a concern is certainly legitimate and underscores the need for a cooperative that is to be successful, to devise an adequate system for quality control and security. A committee responsible for such a system should ·be established at the outset by the sponsor- ing firms.
Protection
One of its duties should be to adequately safe- guard the sponsors' records and documents against fire, theft, water, and other hazards and disasters.
Statistics compiled by the Safe Manufacturers Na- tional Association show that about one-half of com- panies whose important records and documents were destroyed through some catastrophe never resumed business or were permanently closed down within six months; an additional 13% suffered serious economic impairment and were able to remain in business only under severe operating handicaps.5 For this reason the quality control and security committee must see that the cooperative service bureau maintains:
protective devices fireproof vaults
a method of reconstructing any destroyed records
adequate insurance to cover loss of impor- tant client records or documents
The committee must also take measures to prevent one client's records from becoming commingled with another client's records - a not-unlikely occurrence in such an environment.
Preventing Disclosure
Finally, the committee should determine how to prevent disclosure of confidential information. For example, no member of any sponsoring company should be allowed in certain strategic parts of the com- puter center. Likewise, each company's records should be coded and the code number known only to key computer center personnel. In cases requiring utmost security, the actual processing of data should be monitored by a key employee of the com- puter center or possibly by a representative of a
COMPUTERS and AUTOMATION for November, 1973
disinterested third party, such as a firm of inde- pendent public accountants.
Sharing of Expenses: Startup Costs
With an adequate number of sponsoring firms lined up and a sound security system worked out, the third necessity in establishing a successful cooperative is to determine methods of sharing the expenses.
There are two categories of costs to consider:
1. Startup ~osts. These include all costs incur- red prior to the time that the cooperative commences operations. Startup costs are either directly traceable to a particular firm or are joint or com- mon to all firms. An example of a direct cost would be that of analyzing, modifying, and redesigning of systems prior to conversion to the computer. An ex- ample of a joint or common cost would be the cost of computer housing, including air conditioning, engi- neering supervision, false floors, ducts and pipes, tranformers or motor generators, cabling and wiring, and overhead racks and supports.
Each sponsoring firm should pay for its own di- rect costs. Joint costs should be shared equally or apportioned among the firms according to some equi- table formula.
Operating Costs
2. Operating costs. These include the monthly hardware and software costs to operate the coopera- tive. Hardware costs refer to the periodic rental or purchase charge for the computer equipment. As a minimum the computer equipment consists of an input unit, a central processing unit, and an output unit.
Software costs include personnel costs, programming costs, testing and debugging costs, magnetic tapes, disc packs, punched cards, paper, paper tapes, re- pair parts, power, telecommunication lines, and so on.
Operating costs that are directly traceable should be paid by the' using firm. joint or common costs can be shared equally or can be allocated to each firm based on the number of transactions pro- cessed or the actual computer processing time used during the period. In addition, if so desired, a sponsor could be required to pay a minimum or maxi- mum monthly charge.
Management
<'.
Fourthly, the sponsoring companies should decide how to manage the venture. It is advisable to form a committee to oversee the management and operation of the cooperative. A decision must be made whether each member regardless of size should have one vote or whether another basis of voting should be used.
This decision should assure participants that no one firm will dominate or control the cooperative.
Common Line of Business
A factor not to be overlooked when forming a co- operative, one which will avoid frustrations, com- plications, and extra expenses, is to see that each member is in the same industry or business category.
Thus all sponsoring firms will have similar operat- ing problems that can be simultaneously solved by the computer personnel, resulting in the allocation of smaller costs to each firm for each problem solved. Also computer programs developed or pack- aged programs purchased can be shared by all spon- sors with minimum modifications, thus resulting in considerable programming cost savings.
(please turn to page 13)
9
Control in Time-Sharing Systems
Fermin Caro del Castillo 6043 Westridge Lane Fort Worth, Texas 76116
'~ time-sharing installation (/ike every computer environment) is exposed to the seven major dangers: fire, water, theft, fraud, sabotage, equipment malfunctions, and human errors."
Introduction
This article has as its main objective to furnish and define some methods and provisions for control and security in a computer time-sharing environment.
It is evident that computer time-sharing has be- come very popular during recent years. More and more confidential information is being handled by
these systems, creating an urgent need for strong measures of control and security.
Description of the System
The service provided by a single computer to many telecommunications terminals has been called "com- puter time-sharing". With this type of service each user shares simultaneously in the processing capa- bilities of the central processor. This new com- puter facility makes this service available to small companies and/or other users who have little need for their own computer.
Computing services that may go under the heading of time-sharing include:
Commercial computing; text editing; databank information retrieval services; application services such as colleges' records, inventory control, payrolls of small companies, and ac- count receivables; administrative messages;
switching and collection services; and more.
Among the great number of time-sharing users are hospitals (Welch Hospital, one of the biggest in Europe, has acquired the large ICL 1904S computer with 7020 terminals), banks (Gosbank, the national bank of the USSR, has ordered two large-scale Honey- well series 600 with 100 terminals), schools and colleges (British schools and colleges have started using terminals), airlines (Continental Airlines uses the Sonic 360 reservation system with 550 on- line terminals).
Trends
The growth in computer time-sharing has been phe- nomenal in recent years in spite of a short business recession between 1970 and 1972. Time sharing was a broker's dream in the '60s; many companies realized its potential, entered the market, and failed be- cause of: one, strong competitive pressure, and two, because of the countless thefts and violations suf- fered. The highly competitive situation benefited the user in regard to pricing, but the offsetting consequence of lower profits caused a high casualty rate among these companies. As a result, the number of time-sharing firms dropped from 150 to 50 between 1969 and 1972.
10
A few of the companies which survived and which account for most of today's time-sharing business are Rapidata Corporation, Teletype Corporation (a subsidiary of AT&T), Tymeshare Incorporated (the largest independent firm and second only in size to IBM and Honeywell time-sharing operations), and IBM, Jerry Dreyer, executive president of ADAPSO, an association of data processing service organization~
estimates that one time-sharing company out of three was profitable in 1971, two out of five in 1972, and probably three out of six in 1973. Nowadays, this computer facility is coming back to its original trend, and it is expected that the number of com- puters with terminals will grow from 32% at the end of 1971 to 45% at the end of 1975, with the average central processing unit driving 15 to 20 terminals.
One consultant, Creative Strategies of Palo Alto, California, predicts that time-sharing sales by 1976 will increase to 2 billion from 331 million last year. By 1975, says ADAPSO's Dreyer, time-sharing will account for 1/3 of the $4.5 billion computer services industry.
Technological Advances
Most important technological advances have been achieved recently on on-line terminals linked to central processors. Some of the improvements made include solid state keyboards, which have greater reliability and lower cost; visual display methods, which incorporate more capacity, economy, and aes- thetic appeal; non-impact printing techniques, which provide faster, quieter and more reliable opera- tions; improved lower cost memories; more powerful logic capability; and faster and more accurate modern techniques and improvements in central proc- essor software.
This is only the beginning. It is foreseen that in the future the equipment itself will change in nature and will probably not be recognizable as terminals per se. Rather, terminals wiil be modu- lar systems consisting of the required input/output functions for specific jobs, built around basic con- troller and communication interfaces; in many in- stances, terminals will become special purpose de- vices.
Hazards
A time-sharing installation, like every computer environment, is exposed to the seven major dangers of fire, water, theft, fraud, sabotage, EDP equip- ment malfunctions, and human errors. Fire is con-
sidered to be the greatest· threat to magnetic tapes.
Water does not constitute an important hazard to magnetic tapes; but it does to computer installa-
COMPUTERS and AUTOMATION for November. 1973
tions. Theft, fraud, sabotage, EDP equipment mal- functions, and human errors are considered the most common dangers.
On-line terminals connected to central processors from remote points are more exposed to violations and thefts. The increasing popularity of time- sharing systems among large corporations and ser- vice bureaus has given rise to even more potential security breaches. Data transmitted over a com- munication line could be subject to wire tapping and a number of other hazards such as piggyback entry, whereby the intruder intercepts and compromises communication between a terminal and the processor while a legitimate user is inactive but still hold- ing the line open. The intruder can even cancel the user's sign-off signal and continue operating in his name. A knowledgeable person could enter program changes from a terminal and play havoc with the system.
Need for Protection
Due to the increasing popularity of computing services. the issue of control and security pro- tection has become more important. It is evident that time-sharing systems present few obstacles to unauthorized parties. The security problem has been made much more critical by the growing number of people trained in computers and by the fading of the computer mystique. In addition, communication by means of time-sharing systems has no more protection than telephone conversations or Morse-coded methods, since the technological skills necessary to inter- pret computerized data are widespread. More and more companies are appointing security monitors from their EDP staffs to centralize security matters.
Trade organizations such as the American Management Association and the Bank Administration Institute, computer firms, and research firms such as Advance- ment Management Research, Inc. find their seminars on computer security overcrowded by data processing managers and security officers from business and government.
Target of Attacks
Computers have become an important source of in- formation and. as a result, the target of many at- tacks. Some of the general information targets for industrial espionage are sales and service informa- tion, market analysis strategies, bid prices, cor- porate finance, stockholder information, legal ne- gotiations. planned policy changes, expansion plans, product developments, personnel changes, payroll data. general administrative matters; and the list could be expanded even more.
Threats
-The case of an 18 year-old Cincinnati youth who used long distance telephone to tap the lines of a time-sharing system firm in Louisville, Kentucky, and extra<cted data from its ledgers, as well as rec- ords of its customers, is representitive of the type of risk to which computer installations are exposed.
Accidents can have serious consequences such as the incident in which income tax return records for 1969 were erased by energy emitted from the radar of a nearby airport in Austin, Texas.
Safeguards
It is important to consider that too much control and too many security safeguards can become bother- some and costly. The measures to protect data from unauthorized access vary from one system to another.
According to Richard F. Cross, Security Office for the Bank of New York, security systems should in-
COMPUTERS and AUTOMATION for November, 1973
elude each of the following segments: physical·se- curitYi personnel security; procedural security;
audit control; insurancei and any needed interfacing.
The quality and level of protection required de- pends on the sensitivity of the data handled. Never- theless, control and security in a time-sharing en- vironment should encompass the whole system, since it is well known that even the strongest control measures can be violated at the weakest point. These measures should be taken in the central processing unit, software, personnel, communication lines, the terminal, and its users.
Methods and Provisions - Central Processing Unit
The central processor is threatened mostly by sabotage, fire, water, theft, EDP equipment mal- functions, human accidental errors and environmen- tal problems.
The computer center of any company is the heart of the organization, and protection of the equip- ment against these threats can be provided by a carefully planned computer room. Luis Scoma, Presi- dent of Data Processing Security, Inc., recommends that the computer center be located out of the main traffic areas. It must be fireproof, dustfree, and waterproof and provided with temperature and humid- ity control and carbon dioxide fire extinguishers.
Water lines should not run through the computer room. A separate, fireproof storage area should be provided for data files, documentation, and oper- ating supplies. It should also have alarm devices sensitive to magnetism, humidity, heat, pressure, dust, theft, power blackout, etc. There should be strong access control including armed guards, fenced areas, TV monitors, personnel identification (visual, voice print, fingerprint, badge, passwords, etc.), well-performed maintenance service, and a number of extra provisions for protection.
Data Processing Security, Inc. has developed, for high security necessities, an electronically oper- ated double door entry system for access control in- to the computer room. When a person enters the buf- fer zone, the door locks behind him while he is sub- jected to electronic search. If something is de- tected, the system freezes and automatically alerts the security guard. The second door can be opened only with a special badge key.
A carbon dioxide (C02) fire extinguishing system, such as the one installed at the Chase Manhattan Bank's New York City headquarters, has proved use- ful. Engineered by Walter Kidde & Co., Inc., this system uses a battery of remotely located cylinders containing the liquid gas. When activated by smoke detectors, the gas discharges and builds up an in- ert atmosphere in the immediately surrounding area, extinguishing the fire without damaging equipment or data files. Employees can go back to work within five minutes after the fire has been put out and the ventilation systems restarted.
Another fire-fighting system which has been de- veloped by the Ansul Company uses Halon extinguish- ing agents which are discharged in the form of a colorless, odorless, and non-toxic gas. Since Halon agents do not work by diluting oxygen, they are well-suited for areas where humans are present.
The Guaranty Bank and Trust Co. of Worcester, Mass., has adopted another approach to fire-fighting, a system installed by Security Control Systems which detects fire, smoke and excessive temperatures and humidity fluctuations and causes a monitoring and
11
reporting device to alert four pre-selected phone numbers for help.
Software
Protection of data files, whether in the form of punch cards, magnetic tapes, or discs, is the key element in any computer facility.
Some specific examples of software security con- trol measures are audit trails, acce~s regulations, strong supervision of computer operators, the use and control of programs utilizing program documen- tation methods, the use of a log for all significant events (such as user identification, file u~e and attempts of unauthorized use), regulations to pre- vent unauthorized personnel from browsing through the files, and the uS& of a semi-automatic data in- ventory control system.
Personnel
Physical control as to who is allowed near the computer and the files is advisable. Note that this does not ,mean "employees only"; it implies a careful screening of employees, repairmen, and visitors to determine which of them should be allowed this prox- imity to the system. Intimately involved with time- sharing installations are three types of personnel:
operators, programmers, and maintenance engineers.
According to Dennis Van Tassel, mathematician and head programmer at San Jose State College, all oper- ators mu~t und&rstand that there exists a protection philosophy: If personnel are expected to help en- force and to comply with this protection philosophy, it must be clearly defined and specified.
It is advisable to use the following provisions for personnel control: control of logs and monitor, division of responsibilities, rotation of duties, adequate supervision to reduce the risk of losses caused by accident or error, careful handling of data files, cleanliness regulations, control access to private files, and appropriate instruction in case of disasters.
Terminals
Some of the most common pitfalls encountered when using terminals are frequent communication in- terruptions, risk of communication interception, difficult access control, and noisy lines.
Among the greatest disadvantages of time-sharing services are those in connection with the communi- cation lines which link the terminals to the central processing unit. Telephone lines are designed for voice communication rather than data communication, and the results have been that such lines are too noisy and have too many interruptions for appropri- ate handling of data communication.
There have been some recent developments in the improvement of data transmissions. Bell System has developed its Digital Data Service (DDS), a data transmission system which has been introduced early this year in five cities (New York, Boston, Phila- delphia, Washington, and Chicago); and, by the end of 1974, twenty-four cities are expected to have it.
Bell further plans to provide continuous monitoring of their DDS channels. When errors of transmission are detected, DDS will notify the customer. Deci- sion on retransmission will be the customer's re- sponsibility. Another development in data communi- cation is DUV (data under voice). By late 1973 this technique will be in service between New York and Chicago; consequently, communication problems should
12
become virtually nonexistent within the near future, and this system will become even more popular.
Telephone lines are vulnerahle to three types of security problems: wire tapping, piggyback, and user's sign-off signal cancellation. These common threats may be avoided by the use of a privacy transformation method (also called scramblers or cryptographic techniques). This is a non-singular (reversible) operation which conceals the original message either by the substitution of new characters, rearrangement of the characters, or by the adding of strings of digits to the original message.
Some new techniques for encrypting data have evolved, such as high speed transmission and signal scrambling. In spi te of all of these new tech- niques, it is still possible to intercept and tap the information. According to Tuckerman, unauthor- ized users (intruders) who possess only limited ma- terial and information with which to work, can read- ily extract the original text of enciphered mes- sages by making use of the speed, capacity, and computational abili ties of the computer.
The greater the efficiency of the privacy trans- formation, the more difficult it becomes for unauth- orized copying of files.
Access Control
Access to the system throughout the communication lines should be controlled by the central processing unit and by the user's own procedures. This is pos- sible by cryptographic and scramble techniques, per- sonnel regulations, user identification, appropriate terminal location, or by ciphering and deciphering hardware.
Computer software packages have been written to cipher data transmissions. Ciphering software is an attractive technique because it can be done auto- matically by the program which creates the data and at a very low cost. Only programs using the match- ing deciphering technique are able to use this data.
These special hardware cipher devices can be located at the terminal and at the computer to protect transmitted data.
Scramblers or voice privacy devices are currently being used by people who readily acknowledge the in- security of transmitted information. Scramblers are used by large corporations such as the oil industry as well as by union representatives during contract negotiation time. (Scramble phones are currently a 20 to 30 million-dollar business, excluding law en- forcement and military purchases). Scramble devices are now being built for time-sharing operations.
These devices offer protection similar to encryption.
Scramblers convert data into something resembling channel noise or a malfunctioning circuit.
Other features for terminals include print in- hibit, security keylock, operator identification card reader, and the use of a terminal identifier.
With print inhibit the terminal operator can enter data, such as security passwords, without its being physically displayed. A physical keylock can be at- tached to the terminal which will not allow the entry of data. The operator identification card reader reads magnetically encoded information on a magnetic strip card. The card can be used to supply
the user's password and/or name.
Direct access device security features include file mask, volume detection, and write-exhibit switch. The file mask controls or allows read-only
COMPUTERS and AUTOMA nON for November. 1973
or read-write access to data sets. The volume de- tection switch alerts the control program when the volume is increasedj this protects data from being written upon. This switch provides additional pro- tection for read-only volumes.
Insurance
If data security measures fail, the use of EDP insurance can back-up or soften the blow. Every business or person who -now has an in-house computer, operates or is a user of a service bureau, or trans- mits data to a data center should be computer-secur- ity conscious. Insurance is one of the back-up measures if prevention fails. However, not all business can afford or even need an elaborate and expensive security system; thus, evaluation of the information becomes necessary to determine those sections of the system which require this protection.
The subject matter of ordinary insurance coverage is physical damage to tangible property. In this case, software and data are not covered. Business interruption insurance protects you in the event of fire or other interruptions to normal business but does not cover the consequences. Boiler explosion policies which also insure other heating mishaps do not cover computer operations. Theft and employee dishonesty are also insured in terms of general risk. Valuable papers and records' coverage pro- vides protection in the event of destruction of such intangibles as notes and account receivable records but does not include such computer media as valuable papers. Public liability insures the company for its acts or omission of acts which give rise to claims by outsiders against the company.
Insurance can prevent an unfortunate event from becoming a catastrophe. When a person buys an in- surance policy, he knows that he will be indemnified if a loss does occur.
Some of the insurance companies which write EDP policies are the Saint Paul Fire and Marine Insur- ance Co., Insurance Company of North America, The Home Insurance Co., and Royal Assurance of America.
Insurance premiums generally correspond to the regular fire policy rates plus an added loading fac- tor for the additional EDP perils.
Undoubtedly any computer installation, regardless of the quality and level of its security system, cannot approach the state of no risk of loss. Se- curity is based on a "cost-benefit" concept; in other words the cost of violating the system should be greater than the gain derived therefrom.
References
"All About Time-Sharing Services." Data Processing Digest. Vol. 18, No.3, March, 1972.
Bairstow, Jeffrey N. "The Terminal that Thinks for Itself." Computer Decisions. Vol. 5, No.1, January, 1973.
Bartram, Peter. "Software Security." Data Proces- sing Digest. Vol. 18, No.4, April, 1972.
Branstan, Dennis K. "Privacy and Protection in Op- erating Systems." Computer. Vol. 6, No.1, January, 1973.
Bray, Melvyn. "How Safe is Your System?" Dat~
Processing Digest. Vol. 18, No.4, April, 1972.
Chesson, Frederik W. "Computers and Cryptology."
Datamation. Vol. 19, No.1, January, 1973.
Computer News. Vol. 16, No.6, June, 1972.
Computer News. Vol. 16, No. 10, OctOber, 1972.
"Developments in Data Transmission." EDP Analyzer.
Vol. 11, No.3, March, 1973.
COMPUTERS and AUTOMATION for November. 1973
Dickey, C. Lewis. "Securing the Compu ter." Journal of Systems Management. Vol. 23, No.2, February, 1972.
"Distributed Intelligence in Data Communication."
EDP Analyzer. Vol. 11, No.2, February, 1973.
"The Emerging Computer Networks." EDP Analyzer.
Vol. II, No. I, January, 1973.
Girsdansky, M. B. "Cryptology, the Computer, and Data Privacy." Computers and Automation. Vol.
21, No.4, April, 1972.
Goldstein, Seth. "A New Life for Time-Sharing."
Dun's. Vol. 101, No.1, January, 1973.
Hebditch, D. L. "A Simple Terminal Language for a Database." Computer Bulletin. Vol. 16, No.1, January, 1972.
Koehn, Hank E. "Are Companies Bugged about Bugging?"
Journal of Systems Management. Vol. 24, No.1, January, 1973.
Murphey, Wayne E. and Daniel V. Olson. "Controlling , Access to Large Tape Files." Data Processing
Magazine. Vol. 14, No.1, Spring, 1972.
Salzman, Roy M. and Arthur D. Little. "An Outlook for the Terminal Industry in the United States."
Data Processing Digest. Vol. 18, No.1, January, 1972.
"Security in Data Processing." Data Processor.
Vol. 16, No.1, February, 1973.
Simis, T. L. "Telephone Service: The Rules of the Game When the Game is Changing." Computers and Automation. Vol. 21, No. 12, December, 1972.
"Toward Distributed Computing." EDP Industry Re- port. Vol. 7, No.6, January 17, 1972.
Van Tassel, Dennis. Computer Security Management.
Englewood Cliffs, N.J., Prentice-Hall, 1972.
Zaiden, Dennis J. "Special EDP Insurance: Who Needs It?" Data Processing Magazine. Vol. 14, No. I,
Spring, 1972. 0
Cerullo - Continued from page 9
Procedure for Withdrawal
Sixth, a procedure should be established for withdrawal from the cooperative. If the withdraw-
ing firm has not been a member for a predetermined time, perhaps one or two years, a sliding scale penalty fee for withdrawal should be established.
Procedure for Admission
Finally, procedures must be determined to govern the admission of new members into the cooperative.
Perhaps a unanimous affirmative vote by current mem- bers should be required to admit a new member. The current members should also decide if a newly admit- ted member should be required to pay a predetermined portion of the cooperative's startup costs and if the cooperative should limit membership to firms of the same approximate size as current members.
Conclusion
With due care taken in following the above steps, establishing a successful cooperative service bureau should be relatively easy. Companies would be well
advised to consider the idea seriously, for cooper- ative service bureaus offer a viable and in many ways su- perior al ternative to the more costly inhouse computer and the less satisfactory outside service bureau.
Foothotes
1. Computers in Business: An RIA Survey of Users and Nonusers (New York: Research Institute of Ameri- ca, 1969), p. 4.
2. Ibid., p. 17.
3. Michael J. Cerullo. "Service Bureaus: User Ap- praisaL" Datamation, Vol. 18, May 1972, p. 86.
4. Ibid., p. 89.
5. ADAPSO Eighth Management Symposium. "How to Op- erate a Service Bureau for Profits." 1963, p. 49.
o
13
